Segmentation

Remote sensing (RS) aided forest inventory applications are becoming more and more important for visualizing location and amount of forest resources and for providing comprehensive statistics about forests. In case the RS analysis is based on modern very high resolution RS imagery, the traditional pixel-based analysis may not be sufficient but the analysis needs to be carried out using natural image objects rather than single pixels. These objects can be delivered using image segmentation techniques. This paper serves as a brief introduction to image segmentation and its applications in forestry, and is meant for forestry students of different levels. The paper provides basics for understanding what image segmentation is about and how can it be used in solving problems related to forest inventory.

PEKKARINEN Anssi;  HOLOPAINEN Markus; 

2005-10-25

Springer

JRC30652

https://publications.jrc.ec.europa.eu/repository/handle/JRC30652,